Using a similar solution for the callback system (see issue #110) implement a "ping" request to a remote destination so that when an instance fails it can be removed from rotation.
Configurable options should look like:
ping_frequency (in seconds). Note it is critical to allow enough time in between checks on the receiving end to avoid being marked as down.ping_url: the remote host and endpoint that receives the requestping_retries: Retry count for failed ping requests (non 200 responses back)
The code should determine what constitutes a working system for its various systems: database, rabbitmq, http)
Using a similar solution for the callback system (see issue #110) implement a "ping" request to a remote destination so that when an instance fails it can be removed from rotation.
Configurable options should look like:
ping_frequency(in seconds). Note it is critical to allow enough time in between checks on the receiving end to avoid being marked as down.ping_url: the remote host and endpoint that receives the requestping_retries: Retry count for failed ping requests (non 200 responses back)The code should determine what constitutes a working system for its various systems: database, rabbitmq, http)